[clang] Thread Safety Analysis: Support warning on obtaining address of guarded variables (PR #123063)

[Marco Elver via cfe-commits]

melver wrote:

> Doesn't `counted_by` have the same requirements? If not, why does guarded_by?

It does, AFAIK.

> Sure, I can imagine where there might be pushback, but perhaps its worthwhile to see how far you can get with either marking structs that don't need `-fexperimental-late-parse-attributes`, or propose such member declaration re-orderings. At least that would let you land something upstream and begin protecting the kernel/finding bugs.

Fair point, in the worst case individual TUs could still add the flag if needed.

https://github.com/llvm/llvm-project/pull/123063